import os
import tempfile
from pathlib import Path
from typing import Union

# -----------------------------------------------------
# Ensure Streamlit can write its internal files in Spaces
# -----------------------------------------------------
TMP_HOME = tempfile.gettempdir()
# Force HOME to a writable tmp directory (override any existing)
os.environ["HOME"] = TMP_HOME
# Disable anonymous usage stats to avoid metrics file write
os.environ["STREAMLIT_BROWSER_GATHERUSAGESTATS"] = "false"
os.makedirs(os.path.join(TMP_HOME, ".streamlit"), exist_ok=True)
# -----------------------------------------------------
TMP_HOME = tempfile.gettempdir()
os.environ.setdefault("HOME", TMP_HOME)
os.makedirs(os.path.join(TMP_HOME, ".streamlit"), exist_ok=True)

import pandas as pd
import streamlit as st

import pandas as pd
import streamlit as st

# =====================================================
# Configuration & Authentication
# =====================================================

COURSE_PASSWORD = os.getenv("PASSWORD")
if not COURSE_PASSWORD:
    st.error("Course password not configured. Please set the PASSWORD environment variable (key: PASSWORD).")
    st.stop()


def authenticate() -> None:
    """Simple session‑based password gate."""
    if st.session_state.get("authenticated", False):
        return  # already logged in

    pwd = st.text_input("Enter course password:", type="password")
    if st.button("Submit", key="pwd_submit"):
        if pwd == COURSE_PASSWORD:
            st.session_state["authenticated"] = True
            # Rerun so the app shows the rest of the UI
            if hasattr(st, "experimental_rerun"):
                st.experimental_rerun()
            elif hasattr(st, "rerun"):
                st.rerun()
            else:
                st.info("Login successful — please refresh the page.")
                st.stop()
        else:
            st.error("Incorrect password. Please try again.")
            st.stop()

    st.stop()


authenticate()

# =====================================================
# Grading helpers
# =====================================================

def assign_letter(score: float) -> str:
    """Instructor‑specified letter‑grade curve."""
    if score <= 59.99:
        return "F"
    elif score <= 66.99:
        return "D"
    elif score <= 69.99:
        return "D+"
    elif score <= 76.99:
        return "C-"
    elif score <= 79.99:
        return "C"
    elif score <= 82.99:
        return "C+"
    elif score <= 85.99:
        return "B-"
    elif score <= 86.99:
        return "B"
    elif score <= 88.99:
        return "B+"
    elif score <= 90.99:
        return "A-"
    else:
        return "A"

GRADE_ORDER = ["F", "D", "D+", "C-", "C", "C+", "B-", "B", "B+", "A-", "A"]

# =====================================================
# Data loading & preprocessing
# =====================================================

@st.cache_data
def load_data(csv_name: Union[str, Path] = "course_grades_csv.csv"):
    """Load the gradebook, compute quiz averages, course scores, and letters.
    Only students whose final exam is posted are retained.
    Returns: df_final (only students w/ final), quiz_cols list
    """

    csv_name = Path(csv_name)
    search_paths = [
        Path.cwd() / csv_name,
        Path(__file__).resolve().parent / csv_name,
        Path(__file__).resolve().parent.parent / csv_name,
    ]
    file_path = next((p for p in search_paths if p.is_file()), None)
    if file_path is None:
        st.error("CSV file not found. Searched: " + ", ".join(str(p) for p in search_paths))
        st.stop()

    df = pd.read_csv(file_path)

    # Locate quiz columns
    quiz_cols = [c for c in df.columns if c.lower().startswith("quiz")]
    if len(quiz_cols) != 8:
        st.error(f"Expected 8 quiz columns, found {len(quiz_cols)}: {quiz_cols}")
        st.stop()

    # Coerce quizzes to numeric
    df[quiz_cols] = df[quiz_cols].apply(pd.to_numeric, errors="coerce")

    # Quiz average: sort, drop two lowest, then average
    def drop_two_and_avg(row):
        scores = sorted([float(x) for x in row if pd.notna(x)])
        # Remove the first two (lowest) scores
        trimmed = scores[2:] if len(scores) >= 2 else scores
        return sum(trimmed) / len(trimmed) if trimmed else None

    df["Quiz_Avg"] = df[quiz_cols].apply(drop_two_and_avg, axis=1)

    # Keep students whose final is available
    df_final = df[df.get("Is Final Missing?", "") != "Yes, Final Is Pending"].copy()

    # Compute course score
    df_final["Course_Score"] = (
        0.5 * df_final["Quiz_Avg"]
        + 0.2 * df_final["Midterm Score"].astype(float)
        + 0.3 * df_final["Final Score"].astype(float)
        + df_final.get("ExtraCredit", 0).astype(float)
    )

    # Letter grades
    df_final["Letter_Grade"] = df_final["Course_Score"].apply(assign_letter)

    return df_final, quiz_cols


# Load (cached)
df_final, quiz_cols = load_data()

# =====================================================
# User Interface
# =====================================================

st.title("Course Grade Report")

sbuid_input = st.number_input("Enter your student ID (SBUID):", min_value=0, step=1, format="%d")
sbuid = int(sbuid_input) if sbuid_input else None

if sbuid:
    student = df_final[df_final["SBUID"].astype(int) == sbuid]

    if student.empty:
        st.warning("ID not found, or your final exam grade has not been posted yet.")
        st.stop()

    row = student.iloc[0]

    # ---------------- Quiz breakdown ----------------
    st.subheader("Quiz Scores (lowest two dropped)")
    quiz_scores = pd.to_numeric(row[quiz_cols], errors="coerce")
    lowest_two = set(quiz_scores.nsmallest(2).index)

    quiz_df = pd.DataFrame({
        "Score": quiz_scores,
        "Counted in Avg": [col not in lowest_two for col in quiz_cols],
    })
    quiz_df.loc["Average"] = [row["Quiz_Avg"], ""]
    st.table(quiz_df)

    # -------------- Exams & Extra Credit -------------
    st.subheader("Exams & Extra Credit")
    exams_df = pd.DataFrame({
        "Component": ["Midterm Score", "Final Score", "ExtraCredit"],
        "Score": [row["Midterm Score"], row["Final Score"], row.get("ExtraCredit", 0)],
    }).set_index("Component")
    st.table(exams_df)

    # -------------- Course result -------------------
    st.subheader("Your Course Result")
    col_score, col_grade = st.columns(2)
    col_score.metric("Course Score", f"{row['Course_Score']:.1f}")
    col_grade.metric("Letter Grade", row["Letter_Grade"])

    # -------------- Distribution --------------------
    st.subheader("Class‑wide Grade Distribution (students with posted finals)")
    distribution = (
        df_final["Letter_Grade"].value_counts().reindex(GRADE_ORDER).fillna(0).astype(int)
    )
    st.bar_chart(distribution)

# Footer
st.markdown("---")